dfdx VS crates.io

I keep checking the progress on dfdx for this reason. It does what I (and, I assume from context, you) want: Provides static checking of tensor shapes. Which is fantastic. Not quite as much inference as I'd like but I love getting compile-time errors that I forgot to transpose before a matmul.

It depends on the generic_const_exprs feature which is still, to quote, "highly experimental":

https://github.com/rust-lang/rust/issues/76560

Definitely not for production use, but it gives a flavor for where things can head in the medium term, and it's .. it's nice. You could imagine future type support allowing even more inference for some intermediate shapes, of course, but even what it has now is really nice. Like this cute little convnet example:

https://github.com/coreylowman/dfdx/blob/main/examples/night...

And for transformers, it's really early days for dfdx, but it's a library that aims to sit basically at the Pytorch level of abstraction, that the difference is it's not just coded in Rust, but it follows the Rust-y/functional-y philosophy of "if it compiles it runs".

Wow that is super interesting. I actually tried to use GATs at first to be generic over shapes, but I couldn't do it, I'm sure it would be possible in the future though. There is this library dfdx that does something similar to what you mentioned, but it feels a little clumsy to me.

Awesome, I added an issue here https://github.com/coreylowman/dfdx/issues/597. We can discuss more there! The first step will just be adding the device and implementing tensor creation methods for it.

The next release of dfdx includes a CUDA device and implements many ops. The same dev created a new crate, cudarc, for a wrapper around CUDA toolkit.

A question I have is: what are the philosophical/design differences with dfdx? As someone who's played around with dfdx and only skimmed the README of burn, it seems like dfdx leans into Rust's type system/type inference for compile time checking of as much as is possible to check at compile time. I wonder if you've gotten a chance to look at dfdx and would like to outline what you think the differences are. Thanks!

Rust has a rich ecosystem of frameworks and libraries that let you read, parse, and manipulate text files, interact with cloud services and databases, and perform any other job that your project's development workflow may require. And because of its strong typing and tight memory management, you are much less likely to write programs that behave unexpectedly in production.

All the code discussed in this article can be accessed through this repository. For integration with existing projects consider using keyword_extraction crate available on crates.io.

So, be sure to double-check your critical libraries and be sure their alternatives exist in the Rust ecosystem. There’s a good chance the crates you need are available in Rust's crates.io repository.

The previous section was very simple, this section is also very simple but introduces us to cargo which is Rust's package manager, as a JS dev my mind goes straight to NPM.

Now, there has to be a place where all these packages come from. Similar to npmjs registry, where all node packages are registered, stored and retrieved, Rust also has something called crates.io where many helpful packages and dependencies are registered.

Before proceeding, let's check https://crates.io/, the official Rust package registry.

The most thrilling aspect has been the joy of writing the backend. It's like every struct, enum, and method in Rust forms this interconnected Multiverse of code , which you can see in crates.io which is best Documentation experience I Ever Had.

5. Crates.io

When using crates from eg. crates.io, and also things like std and core

Vendoring your packages was very tedious to even remotely get to work with Cargo. I spent a very long time getting Cargo to work together with cargo-local-registry. We vendor crates from crates.io and a custom internal registry.